This project is coordinated with P50 ES012383, Nicole Lurie, ?UNDERSTANDING NEIGHBORHOOD IMPACTS ON HEALTH.? This project investigates how contextual factors, specifically the built and socioeconomic features of neighborhoods, affect the health of elderly Americans. Neighborhood characteristics may influence the health of elderly persons in two principal ways: through direct contemporaneous effects on health status and functioning, or through cumulative effects from earlier points in the lifecourse. Because many neighborhood features are modifiable, identifying changes in those aspects of neighborhood environment with the greatest influence on health could become important in addressing health disparities and improving health overall. Understanding at what points in the lifecourse the neighborhood environment most affects health will further inform interventions to reduce disparities in health and functioning in old age. Unfortunately, there is a paucity of literature on how these contextual factors affect late-life health and functioning. The few existing studies of the elderly are based on small sample sizes drawn largely from confined geographic areas and are thus of limited generalizability. We propose to undertake a national study of the effect of neighborhoods, both built and social environments on health and functioning in old age. The study has the following aims: 1) to measure the effects of the built and social environment on the health and health trajectories of the elderly, with a focus on identifying those neighborhood features that have the most impact on health and are amenable to policy interventions; 2) to understand how and at what points neighborhood environments from earlier in life influence health in old age; and 3) to examine whether neighborhood features have differential impacts on important subpopulations of the elderly, such as men and women, people with low socioeconomic status, and racial/ethnic minorities. The study will use two nationally representative panel data sets, the Health and Retirement Survey (1992-2002) and the Panel Study of Income Dynamics (1968-2003). These data sets contain large samples of elderly persons that are representative of the elderly population of the United States. Neighborhood characteristics, derived from the work of the Center's data core, will be merged with the individual level data at the census tract level.